45 deg Z printers - conveyor belt ideas

As 45 degree conveyor printers are becoming available commercially (see BlackBelt 3D and PrintrBelt), I have developed an interest in replicating the concept in my RepRap.

The frame and XY mechanics (Core XY) are easy enough, and the first layer adhesion problem is solvable (nozzle camber, and making sure the first 'Z' layer is extruded thickly enough), but the tricky bit from a reprap perspective is the conveyor.

The commercial versions seem to be either (a) a proprietary surface belt (BlackBelt 3D) or (b) a stainless steel belt covered with Kapton tape.

I've looked into ordering a bespoke stainless belt from Misumi USA ($700 for 200mm x 800m) - the smallest ring size) and was put off by the price.

So, I've started looking at what's on the shelf of my local hardware store.

Which led me to: mini bandsaw blades.

They are relatively thin, but in the correct ring size range for a small reprap. I'm thinking of using a set of 10 or so around two recycled printer rollers, mounted flush to each other, then covering them with kapton tape.

10 can be purchased for approx $50 USD in the smaller ring sizes, suitable for experimentation.

Between the rollers, I'd place a standard printer bed heater, in direct contact with the bandsaw blades. Bandsaw teeth would, of course, be arranged such that they are not cutting into the heating plate.

For those of you mechanically inclined, what I am missing?

* Tension springs on the non-drive pulley, or should I use adjustment screws?
* Would the printer rollers (rubber to steel friction, and a bend radius of approx 10mm) be sufficient for this purpose?
* Gearing for the drive stepper may be a concern, and the conveyor will probably be relatively difficult to turn (especially when fully loaded with plastic).
* Will the surface tension of the Kapton be enough to compensate for the gap between teeth of one band and the edge of the next?
* Do I need to cut a groove in the roller to hold the bands, to keep them from peeling off the edge of the roller?
* All the things I'm missing....

The Kapton tape I'm using doesn't stick well to most surfaces.
The teeth of the bandsaw might perforate the tape and make it fail.

Why not using belts of belt sanders instead? You'd have to turn them inside out, but the paper is a good print surface.
The ( fine ) grid will help keeping the belt(s) from slipping.
Covered with Kapton tape the paper would be reusable. ( bare paper: not so much )

Thank you! That’s an excellent idea - I had gotten so hung up on the steel belt I forgot to think of simpler alternatives!

Look at Dorner conveyor belts. Plenty on eBay.
Either the Urethane or PVC belt may work well if coated with Garolite film or Kapton tape.

Keeping a flat surface seems to be very difficult. The Blackbelt has been well designed but still seems to show some banding in the prints most likely due to issues with unwanted movement on the belt system. The printrbot alternative looks rushed and the prints have even worse banding. I'm very surprised they havn't had backlash from the community for blatantly ripping off Blackbelt, but people seem happy just because it's a lower price.

There is a reason they have invested heavily in the design of the Carbon fibre belt. Low stretch, easy to tension, relatively easy to produce in a continuous loop and I imagine the surface provides excellent surface adhesion. If you want a production line or long prints the surface adhesion will play a very important role. Imagine coming back to the printer the next day to see a failed print and a few kgs of wasted plastic?

Quote
Origamib
I'm very surprised they havn't had backlash from the community for blatantly ripping off Blackbelt, but people seem happy just because it's a lower price.

You should do some research before saying things like this. Polar 3D had working prototypes of this back in early 2016 at trade shows.

Quote
Feign

Quote
Origamib
I'm very surprised they havn't had backlash from the community for blatantly ripping off Blackbelt, but people seem happy just because it's a lower price.

You should do some research before saying things like this. Polar 3D had working prototypes of this back in early 2016 at trade shows.

Perhaps I may have been hasty to say they are ripping them off, and will withdraw my comments but Blackbelt have been developing their tech for over 3 years. I find it convenient that printrbot just released their version during blackbelts Kickstarter campaign. I believe the research shows in the Blackbelt design though, they have some pretty impressive results.

Just to be clear, I’m not trying to rip off either Polar/Printrbelt nor Blackbelt.

Primarily this is for my own entertainment, and to research/develop 45 degree slicing technology.

My first prototype is actually going to be a standard core XY platform, with a standard reprap slide-on-Y a 45 degree angle to simulate a conveyor (for developing the software).

Quote
Ezrec
Just to be clear, I’m not trying to rip off either Polar/Printrbelt nor Blackbelt.

Primarily this is for my own entertainment, and to research/develop 45 degree slicing technology.

My first prototype is actually going to be a standard core XY platform, with a standard reprap slide-on-Y a 45 degree angle to simulate a conveyor (for developing the software).

Sorry the comments were not aimed at you. Just my observation that belt material will be highly important and Blackbelt have really cracked that problem it seems. If it can be tensioned well it will slide over a flat surface evenly and give better layer registration. The Kapton alternatives not so much.

Quote
Origamib
I find it convenient that printrbot just released their version during blackbelts Kickstarter campaign. I believe the research shows in the Blackbelt design though, they have some pretty impressive results.

I don't think it's a coincidence at all, but not due to any kind of copying. Here's how I think the timeline went (speculation ahead).
-Bill Steele (of Polar3D) had his conveyor belt printers at several trade shows, but didn't seem to think there would be a market for his desktop-sized printer, thinking it was more of an industrial application, and his company's image wasn't really right for a Cartesian printer. (he also had a DLP-SLA conveyor belt printer at the show I went to, which blew my friggin mind, but apparently the calibration/slicing for it is still more an art than a science at this point.)
-Blackbelt's KS teaser is a huge hype machine (and rightfully so) which gets lots of press.
-Steele knows that if he doesn't make his development more widely known, it might get patented out from under him, so he makes his design public on his Youtube and on Thingiverse and points to research by himself and others that goes back 5 years.
Here's where I go entierly into speculation:
-When he made his design public, he probably still didn't have any intent to commercialize it as a Polar3D product, but Printrbot talked him into partnering with them.
-The Printrbelt is just a mass-production of the design that Steele has already had printing for some time, so it was very quick to get into production.

I just thought of something interesting: What if instead of using a seamless conveyor belt made out of premium materials, the printer feeds in a series of flat aluminum or plastic "build bars" that press tightly against each other as they feed through the machine to act as one large build plate?

Interesting alternative!
The "bars" could even have the size of a normal print bed. They are stacked at the back of the printer. That'd work for overnight printing.

I was thinking bars so they could be separated from the print one bar at a time to reduce the stress on the print from peeling them off.

Here's a completely different idea: Turn the whole thing around, so the printer is on wheels and slowly rolls along a flat floor, printing a trail of parts behind it.

Printer on wheels? Sounds like the idea I had a few month back for a laser engraver, which would allow to engrave long wooden planks.

Its not a new idea... and still suffers from same issues as last time..

No way to get accurate positional information easily.

Would probably need some sort of grid on the floor and lots of cameras, to work out location, and would need to be real time for extrudes. (cant have it drifting of centre...)

I thought of rack and pinion combo, where the pinions are stepper driven wheels. The rack could be something simple, like a slotted metal strip or even toothed belt fixed on a straight extrusion.

Quote
Dust
Its not a new idea... and still suffers from same issues as last time..

No way to get accurate positional information easily.

Not exactly, other rolling 3D printers would travel large distances on each layer, and would be rolling over the lower layers of their own printing if printing something larger than the printer's own wheel base.
This concept would only step the wheels 1/3rd of a millimeter or so for each layer, making its movement across a floor very very slow, and giving it plenty of time to precisely adjust its course.

Quote
Feign
I was thinking bars so they could be separated from the print one bar at a time to reduce the stress on the print from peeling them off.

Here's a completely different idea: Turn the whole thing around, so the printer is on wheels and slowly rolls along a flat floor, printing a trail of parts behind it.

The scatological references alone are worth the research!

I had a part the other day I thought might be better if I printed it at 45 degrees on standard cartesian, ie rotate the model in Z, I would have done it if I had some Printbite or something, still wanted to do it to see if I could manage it as bed is sticking quite well at the moment, so wouldnt have needed special slicing software.

Quote
MechaBits
I had a part the other day I thought might be better if I printed it at 45 degrees on standard cartesian, ie rotate the model in Z, I would have done it if I had some Printbite or something, still wanted to do it to see if I could manage it as bed is sticking quite well at the moment, so wouldnt have needed special slicing software.

I've actually run into that situation quite a few times Just as often, I've printed with a part anchored on a big edge chamfer in order to keep the layer orientation from lining up with shear stresses.

Yeah thats what this part was, on a chamfered edge, but it was about 40x10mm footprint and overhang would have extended way outside this, I hope to try it at some point, but defo do-able with some build surface.

Great initiatives.
I follow with pleasure.

I just noticed Blackbelt has released their version of Cura that includes angled slicing: Blackbelt Cura